Acidic pH as a physiological regulator of human cathepsin L activity

Eur. J. Biochem. 259, 926±932 (1999) q FEBS 1999

Acidic pH as a physiological regulator of human cathepsin L activity

Boris Turk1,2, Iztok Dolenc1, Brigita LenarcÏicÏ1, Igor KrizÏaj1, Vito Turk1, Joseph G. Bieth3 and Ingemar BjoÈrk2

1Department of Biochemistry and Molecular Biology, J. Stefan Institute, Ljubljana, Slovenia; 2Department of Veterinary Medical Chemistry,

Biomedical Center, Uppsala, Sweden; 3Laboratoire d'Enzymologie, Universite Louis Pasteur de Strasbourg, France

Human cysteine protease cathepsin L was inactivated at acid pH by a first-order process. The inactivation rate

decreased with increasing concentrations of a small synthetic substrate, suggesting that substrates stabilize the active

conformation. The substrate-independent inactivation rate constant increased with organic solvent content of the

buffer, consistent with internal hydrophobic interactions, disrupted by the organic solvent, also stabilizing the

enzyme. Circular dichroism showed that the inactivation is accompanied by large structural changes, a decrease in

a-helix content being especially pronounced. The high activation energy of the reaction at pH 3.0 (200 kJ´mol21)

supported such a major conformational change occurring. The acid inactivation of cathepsin L was irreversible,

consistent with the propeptide being needed for proper folding of the enzyme. Aspartic protease cathepsin D was

shown to cleave denatured, but not active cathepsin L, suggesting a potential mechanism for in-vivo regulation and

turnover of cathepsin L inside lysosomes.

Keywords: cathepsin L, cysteine proteinase, cathepsin D, regulation.

Cathepsin L is a lysosomal cysteine proteinase, whose mainfunction is the degradation of proteins in lysosomes [1]. It hasbeen shown to be the most active lysosomal proteinase in thein-vitro degradation of several protein substrates, such asazocasein, collagen or elastin [2±4]. Cathepsin L has alsobeen found to be secreted outside the cell, either in its matureform or as a proenzyme [5], although the extracellular role ofcathepsin L is not yet clear. However, there are severalreports describing the involvement of such secreted forms inpathological processes [6±8]. In addition, the ability ofcathepsin L to degrade matrix proteins is well documented[9; reviewed in 1].

Cathepsin L is ubiquitous in mammalian cells, and all speciesvariants studied have similar properties [10±15]. The aminoacid sequences of human, rat, mouse, sheep and chickencathepsin L are known [14±18] and recently three-dimensionalstructures of procathepsin L [19] and cathepsin L/E-64 complexhave been reported [20]. Like other papain-like proteinases,cathepsin L is synthesized as an inactive proenzyme, which isprocessed to the mature enzyme by proteolytic removal of apropeptide [18,21]. Cathepsin L is optimally active at slightlyacidic pH and in a reducing environment, as are other lysosomalcysteine proteinases [2]. However, mature cathepsin L is veryunstable under neutral or slightly alkaline conditions [13,22±24],because of irreversible denaturation of the protein [23,25]. Incontrast, the precursor form is substantially more stable undersuch conditions [5]. Small synthetic substrates have been shownto have a stabilizing effect on cathepsin L, attributed to a

substantially higher stability of the enzyme-substrate complexthan of the enzyme alone [23]. A general ligand-stabilizatingeffect has been supported by the finding that endogenous proteininhibitors, cystatins, also stabilize the enzyme. Active enzymewas released from such a complex with a cystatin even afterprolonged incubation under neutral conditions [23], a findingthat was repeated with another, nonhomologous competitiveprotein inhibitor, p41 invariant chain fragment [26,27]. Incontrast, very little is known about the stability of cathepsin L inacidic medium, except an incidental observation by Mason et al.[13] that the enzyme is unstable below pH 4.0.

Cathepsin D is the major lysosomal aspartic proteinase and iswidely distributed in almost all mammalian cells [28]. Togetherwith cathepsin B it is the most abundant of the lysosomalproteinases, although the concentration of cathepsin L can be ina similar range (<1 mm in the lysosomes [29]). Cathepsin D isconsidered to be involved in physiological processes such asintracellular protein catabolism [30] and antigen presentation[31], and also in a number of pathological conditions [32].However, experimental studies with cathepsin D-deficient micehave suggested that its major role in intracellular proteincatabolism is not protein degradation inside lysosomes, whichcould be compensated by cysteine proteinases, but rather acti-vation and/or inactivation of biologically active proteins bylimited proteolysis [33].

In this work, we have characterized the irreversible acidinactivation of human cathepsin L, which together with proteo-lytic degradation of the denatured enzyme by cathepsin D, couldbe an important mechanism for in-vivo regulation and turnoverof cathepsin L, and other cysteine proteinases, inside lysosomes.

MATERIALS AND METHODS

Z-Phe-Arg-pNA was from Bachem (Bubendorf, Switzerland),dithiothreitol and dimethylsulfoxide from Aldrich (Milwaukee,WI, USA), EDTA and dimethylformamide from Sigma (StLouis, MO, USA) and N-methylpyrrolidone from Fluka Chemie(Buchs, Switzerland). E-64 was from Peptide Research Institute

Correspondence to B. Turk, Department of Biochemistry and Molecular

Biology, J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia.

Tel.: + 386 61 1773772, Fax: + 386 61 273594, E-mail: [email protected]

Abbreviations: E-64, l-3-carboxy-trans-2,3-epoxypropyl-leucylamido-(4-

guanidino) butane; pNA, p-nitroanilide; Z-, benzyloxycarbonyl.

Enzymes: cathepsin L (3.4.22.15), cathepsin B (3.4.22.1), papain (3.4.22.2),

cathepsin D (3.4.23.5)

(Received 5 August 1998, revised 20 November 1998, accepted 24

November 1998)

q FEBS 1999 pH regulation of cathepsin L (Eur. J. Biochem. 259) 927

(Osaka, Japan). All other chemicals used were of analyticalgrade. Human cathepsin L was isolated and active-site titratedwith E-64 as described previously [23,34]. Bovine cathepsin Dwas isolated as described previously [35]. Concentrations ofcathepsin L were obtained by measurements of absorbance at280 nm with the use of a molar absorption coefficient of44 300 m21 cm21, calculated from the amino acid sequence[16] by the method of Pace et al. [36]. Cathepsin D concen-trations were measured by the method of Lowry et al. [37] withbovine serum albumin as standard.

Buffers

The buffer used for the pH range 3.0±3.75 was 150 mm sodiumcitrate, containing 1 mm EDTA, whereas 100 mm sodiumacetate, containing 50 mm NaCl and 1 mm EDTA, was used atpH 4.0 and 4.2. The activating buffer was 50 mm sodium acetate,pH 5.5, containing 1 mm EDTA and 5 mm dithiothreitol.

Kinetics of cathepsin L inactivation

Cathepsin L was activated for 5 min in the activating buffer, and30 mL of activated enzyme (30 nm final concentration) was thenadded to 970 mL of buffer of low pH, prewarmed to the desiredtemperature (37 8C, except in studies of temperature depen-dence). After appropriate times, 50 mL aliquots were taken foranalysis of residual enzyme activity and added to 950 mL of50 mm acetate buffer, pH 5.5, containing 1 mm EDTA, with100 mm Z-Phe-Arg-pNA as a substrate. The release ofp-nitroaniline was monitored continuously at 410 nm in aUvikon 941 spectrophotometer (Kontron, Milano, Italy) for10±60 s. Usually, 10±15 data pairs were obtained for eachinactivation curve.

The progress of the reactions was also monitored continu-ously in the Uvikon 941 spectrophotometer as follows: 10 mL ofactivated cathepsin L (1.8±20 nm final concentration) wasadded to 990 mL of prewarmed (37 8C) inactivating buffer,containing the substrate Z-Phe-Arg-pNA at various concentra-tions, and the release of p-nitroaniline was then observed at410 nm.

Fast reactions were analysed in a DX-17 MV stopped-flowapparatus (Applied Photophysics, Leatherhead, UK), as des-cribed previously [23].

Reversibility of the acidic pH inactivation of cathepsin L

The reversibility was studied as described previously for thealkaline pH-induced inactivation of cathepsin L [23], except thatthe enzyme was inactivated for 30 min at pH 3.0 or for 8 or 20 hat pH 4.0, and that the reversibility was followed for 48 h. Thereversibility was also studied by another method: 50 mL ofcathepsin L (30 nm final concentration), inactivated at pH 3.0,was added to 950 mL of prewarmed (37 8C) 300 mm acetatebuffer, pH 5.5, containing 1 mm EDTA, 2 mm dithiothreitol,3% (v/v) dimethylsulfoxide and 100 mm Z-Phe-Arg-pNA. Therelease of product was then monitored continuously for 2 h.

Circular dichroism measurements

Circular dichroism was measured in the far-UV region(250±200 nm) at 25 8C) with an AVIV 62 A DS circulardichroism spectrometer (Lakewood, NJ, USA). A cell with apathlength of 1 mm was used. Measurements at pH 5.5 weremade using 50 mm acetate buffer, containing 1 mm EDTA, at aprotein concentration of 0.1 mg´mL21. After the spectrum hadbeen recorded at this pH, HCl was added in a negligible volume

until pH 4.0 or 3.0 was obtained. This was followed byincubation for 24 h (pH 4.0) or 2 h (pH 3.0) at 37 8C, and thespectrum was then remeasured. Mean residue ellipticities werecalculated from a mean residue weight of 112. Time-dependentdenaturation of cathepsin L was followed at pH 3.0 and 30 8C ina 1-mm cuvette as follows: preactivated cathepsin L (finalconcentration 0.07 mg´mL21) was added to the prewarmedinactivation buffer (150 mm citrate, containing 1 mm EDTA)and the decrease in ellipticity continuously monitored at222 nm.

Proteolytic degradation of cathepsin L by cathepsin D

Cathepsin L (0.7 mm final concentration) was inactivated for12 h at 37 8C in 1 mL of 150 mm citrate buffer, pH 3.8, prior tothe addition of 50 mL of cathepsin D (0.35 mm final concentra-tion). The mixture was incubated for 1 h at 37 8C, followed byseparation on HPLC (see below). In a control experiment,cathepsins L and D were incubated alone for 13 h and 1 h,respectively, under the same conditions. The two proteins werethen mixed immediately before loading on the HPLC column.

Peptide separation, purification and N-terminal amino acidsequencing

Peptides were separated by HPLC (Milton Roy, FL, USA) on areverse phase Vydac C18 column (The Separations group, CA,USA), equilibrated with 0.1% (v/v) trifluoroacetic acid. Thecolumn was eluted with an acetonitrile gradient (0±65%acetonitrile) at a flow rate 1 mL´min21, and the absorbance ofthe effluent was monitored at 215 nm. Automated sequenceanalyses were performed in an Applied Biosystems, Foster City,CA, USA, 492 Protein sequencer.

RESULTS

Inactivation kinetics of cathepsin L

Exposure of cathepsin L to pH 3.0 resulted in a progressivedecrease of enzyme activity, monitored by a discontinuous

Fig. 1. Progress curve for the inactivation of cathepsin L in the presence

of substrate at pH 3.0 and 37 8C. The reaction mixture contained 9 nm

cathepsin L and 100 mm Z-Phe-Arg-pNA. Other experimental details are

described in the Materials and Methods section. The experimental curve and

the theoretical curve (based on a first-order reaction, calculated using

Eqn (1) and the best estimates of kobs and vsd obtained by nonlinear

regression analysis) are indistinguishable.

928 B. Turk et al. (Eur. J. Biochem. 259) q FEBS 1999

assay, that was best described by single exponential decay (notshown). As no remaining activity could be detected at the end ofthe reaction, the rate constant was obtained by fitting the data toa simple first-order equation and was found to be 0.0067 s21.Continuous monitoring of activity in the presence of substrate(Z-Phe-Arg-pNA at various concentrations) at pH 3.0 resultedsimilarly in an exponential inactivation, reflected in an expo-nential production of pNA. The substrate consumption was lessthan 3%, showing that enzyme inactivation, rather than dis-appearance of substrate, was responsible for the loss of activity.For this procedure, a small residual activity, which could not beneglected, was observed, and the progress curves obtained(Fig. 1) were therefore best fitted to the following modifiedfirst-order Eqn (1) [38].

P � P1´�1 ± e ± kobs´t� � ksd´t �1�

where P represents the product concentration at a given time,P1 represents the product concentration at infinite time for thereaction in the absence of the residual, steady-state, activity, kobs

is the observed first-order rate constant, and ksd is the rate of thesteady-state substrate hydrolysis. This residual activity ofcathepsin L probably resulted from adsorption of the proteinon the walls of the cuvette, as no such residual activity wasdetected by the discontinuous method. The observed inactiva-tion rate constant was found to be independent of the enzymeconcentration in the range 1.8 to 15 nm, indicating that autolysiswas not responsible for the loss of activity. By contrast, kobs

decreased with increasing substrate concentration (Fig. 2), asalso observed at neutral pH [23]. This dependence is consistentwith a competitive process, in which the enzyme cannot beinactivated when it is in complex with substrate, as shown inScheme 1 [23].

According to this scheme, kobs is given by the followingrelationship Eqn (2):

kobs � kinac/�1 � So/Km� �2�

where So is the substrate concentration. As shown in Figure 2,the experimental data could be satisfactorily fitted to thisequation by nonlinear regression analysis, giving the bestestimates of 200 ^ 20 mm for Km and 0.012 ^ 0.004 s21 forkinac.

Effect of organic solvents on the inactivation rate constant

The value of kinac calculated from Eqn (2) (0.012 ^ 0.004 s21)was higher than that obtained by the discontinuous method(0.0067 ^ 0.0003 s21), indicating that dimethylsulfoxide inwhich the stock solution of the substrate was prepared may havea destabilizing effect on cathepsin L even at the low concen-trations used (3%). We therefore studied the dependence of kobs

on dimethylsulfoxide concentration in citrate buffer, pH 3.0,containing 200 mm Z-Phe-Arg-pNA as substrate. The value ofkobs increased almost linearly with organic solvent concentrationin the range 3±15% (v/v) (Fig. 3). Because this effect might be aresult of the organic solvent affecting both kinac and the Km

value for the substrate (see Eqn 2), we also studied the substratedependence of kobs at 15% (v/v) of dimethylsulfoxide, thehighest concentration used. A similar stabilizing effect of thesubstrate as at 3% (v/v) of dimethylsulfoxide was observed (notshown), the data giving Km and kinac values of 260 ^ 24 mm and0.024 ^ 0.001 s21, respectively, by nonlinear regression fittingto Eqn (2). The Km values determined at 3 and 15% (v/v) of

Fig. 2. Effect of substrate concentration on the observed rate constant

for cathepsin L inactivation at pH 3.0 and 37 8C. The observed

inactivation rate constant, kobs, was determined as described in the Results

section, using 2 nm cathepsin L and variable concentrations of Z-Phe-Arg-

pNA. Each point represents an average of at least 3 different experiments at

the same substrate concentration, the bars giving the standard errors. The

solid line is the theoretical curve, calculated using Eqn (2) and the best

estimates of kinac and Km obtained by nonlinear regression analysis.

Scheme 1. Mechanism of enzyme inactivation in the presence of

substrate. E and E* denote active and inactive cathepsin L, respectively, S

and P represent substrate and product, respectively, and kinac is the first-order

inactivation rate constant.

Fig. 3. Effect of dimethylsulfoxide concentration on the observed rate

constant for inactivation of cathepsin L at pH 3.0 and 37 8C. The

observed rate constant, kobs, was determined as described in the legend to

Fig. 2. The experiments were performed at least in triplicate, the bars giving

the standard errors. Other experimental details were as described in the

Materials and Methods section. The solid line was generated by linear

regression analysis.

q FEBS 1999 pH regulation of cathepsin L (Eur. J. Biochem. 259) 929

dimethylsulfoxide did not differ significantly (200 ^ 20 mmand 260 ^ 24 mm, respectively), indicating that the organicsolvent predominantly affected kinac. In additional experi-ments, the substrate was dissolved in dimethylformamide andN-methylpyrrolidone, solvents with higher hydrophobicicitythan dimethylsufloxide, resulting in substantial increases ofthe observed inactivation rate constant in the presence of200 mm Z-Phe-Arg-pNA and 3% organic solvent, to 0.017 ^0.001 s21 and 0.033 ^ 0.001 s21, respectively.

pH and temperature dependence of the inactivation rateconstant

The rate of inactivation of cathepsin L was studied in the pHrange 3.0±4.2 by the discontinuous method. The inactivation

rate constant increased exponentially with decreasing pH to avalue of 6.7 £ 10±3 s21 at pH 3.0. From a plot of log kinac vs.pH (Fig. 4), one proton was found to be adsorbed in pH range3.5±4.2 and three were adsorbed in the pH range 3.0±3.5,indicating that two groups with pKa values <3.5 (intercept ofthe two lines in Fig. 4) are involved in the rate limiting steps ofthe inactivation process.

The inactivation rate constant was also affected by tempera-ture. An activation energy of about 200 kJ´mol21 was calculatedfrom an Arrhenius plot (Fig. 5) of measurements between 15and 37 8C.

Structural studies of cathepsin L by circular dichroism

The far-UV circular dichroism spectra of cathepsin L at pH 5.5,4.0 and 3.0 (Fig. 6) indicates that practically all of the secondarystructure was lost after exposure of the enzyme to pH 3.0 orpH 4.0. This observation is consistent with the acid inactivationof cathepsin L being accompanied by a substantial unfolding ofthe protein. The structural changes accompanying cathepsin Ldenaturation were monitored by continuous measurements ofcircular dichroism at 222 nm at 30 8C. The progress curves weresingle exponentials (not shown), indicating that unfolding ofcathepsin L is a simple, unimolecular process. The value of theunfolding rate constant (kunf = 9.0 ^ 1.0 £ 10±4 s21) was inreasonable agreement with the value for the inactivation rateconstant determined under the same conditions (kinac =13.3 ^ 1.5 £ 10±4 s21), indicating that the two processes arelinked. This is in agreement with previous studies on cathepsinB [38].

Reversibility of inactivation

Reversibility was studied using the discontinuous method afterinactivation at pH 3. Between 7% and 8% of the originalactivity was regained in the first 24 h, and this activity remainedconstant for the next 24 h. Similarly, continuous monitoring ofreversibility in the presence of substrate showed a rapidrecovery of activity to about 5% of the control, followed by aconstant substrate degradation for up to 2 h (the maximum timeof continuous observation). Variation of the inactivation timebetween 10 and 60 min did not result in an increased recovery

Fig. 4. pH-dependence of the inactivation rate constant of human

cathepsin L at 37 8C. The inactivation rate constants were measured by a

discontinuous assay. Experimental conditions were as described under

Materials and Methods. By linear regression analyses slopes of (3.0 ^ 0.1)

and (0.96 ^ 0.07) were calculated in the pH ranges 3.0±3.5 and 3.5±4.2,

respectively.

Fig. 5. Arrhenius plot for the first-order inactivation rate constant at

pH 3.0. The inactivation rate constant, kinac, was determined by the

discontinuous method (see Materials and Methods). T is absolute

temperature. The solid line was generated by linear regression analysis,

the slope giving the activation energy.

Fig. 6. Far-UV circular dichroism spectra of cathepsin L at pH 5.5

(ÐÐ), pH 4.0 (± ± ±) and pH 3.0 (´ ´ ´ ´). The concentration of cathepsin

L was 0.1 mg´mL±1. Other experimental details were as described in the

Materials and Methods section.

930 B. Turk et al. (Eur. J. Biochem. 259) q FEBS 1999

of activity. Similar results were obtained at pH 4.0, suggestingthat the inactivation of cathepsin L at acidic pH is essentially anirreversible process.

Effect of cathepsin D

In an initial experiment, cathepsin L (30 nm final concentration)was incubated in the absence and presence of cathepsin D(0.1 mm final concentration) at pH 3.5 or 4.0 and 37 8C asdescribed in Materials and Methods. The inactivation rateconstants for cathepsin L inactivation in the absence or presenceof cathepsin D, measured by the discontinuous method, wereindistinguishable. This indicated that cathepsin D did not cleaveactive cathepsin L or that the cleavage rate was substantiallylower than that of the spontaneous pH-inactivation. Incubationof denatured cathepsin L with cathepsin D at acidic pH (3.8),however, resulted in partial cathepsin L degradation. The result-ing peptides were separated and the N-terminal sequences of themain degradation products were determined to locate the cleav-age sites in cathepsin L. The principal and most extensivecleavage occurred between Cys98 and Lys99, which was fol-lowed by further degradation of the molecule at other sites, e.g.at Gly149-Ile150 and Gly191-Met192 (Fig. 7).

DISCUSSION

Cathepsin L was found to be irreversibly inactivated at acid pH.The process was accompanied by a considerable loss of struc-ture, as demonstrated by circular dichroism measurements atpH 3.0 and 4.0, the decrease of helical content being especiallypronounced. These findings are typical for an acid-denaturedprotein [39]. The conclusion that inactivation involves a majorstructural change is supported by the high activation energy ofthe process (200 kJ´mol21), which is comparable with values forthe alkaline inactivation of cathepsin L (175 kJ´mol21 [23]) andcathepsin B (184 kJ´mol21 [38]). Moreover, the irreversibilityof the inactivation process is consistent with the propeptidebeing needed for proper folding of cathepsin L [40].

Kinetically, the acid inactivation of cathepsin L was a first-order process, as also observed for other related cysteineproteinases at both acid and alkaline pH [23, 38] (B. Turk and V.Stoka, unpublished results). The rate of the process decreasedsubstantially with increasing concentrations of a small syntheticsubstrate, as was also observed for the inactivation at neutralpH [23]. This finding is probably a result of a competitivemechanism, the results strongly suggesting that the enzyme

cannot be inactivated in complex with the substrate, as proposedpreviously [23]. However, the Km values for the hydrolysis ofthe substrate, Z-Phe-Arg-pNA, at acid pH were substantiallyhigher than those observed at neutral pH [25], reflectingappreciably weaker substrate binding and resulting in a smallerprotective effect of the substrate at acidic pH. The substrate-independent inactivation rate constant increased in the presenceof dimethylsulfoxide, indicating the importance of internalhydrophobic interactions disrupted by the organic solvent, forthe stability of cathepsin L. This effect was even more pro-nounced with two, more hydrophobic solvents, dimethylforma-mide and N-methylpyrrolidone [41,42], supporting this idea.Because a similar destabilizing effect was observed also forcathepsin B [38,43], this may be a general phenomenon for thepapain-like cysteine proteinases. Moreover, these results indi-cate that the values of the substrate-independent inactivationrate constants obtained in a previous study of cathepsin Linactivation at neutral pH were too high, as 5% (v/v) dimethyl-formamide was used in these experiments [23].

The inactivation of cathepsin L was found to be stronglydependent on pH. The inactivation rate constant increasedexponentially with decreasing pH, indicating that the pH-induced inactivation of cathepsin L is catalyzed by protons. Thisfinding also suggests that ionic interactions are important for thestability of cathepsin L and other cysteine proteinases, as pre-viously proposed from inactivation studies at neutral pH[24,25,38,44]. Moreover, detailed studies on buffer compositionhave indicated that different ions have different effects onstability of cysteine proteinases at neutral pH [24,44].

On the basis of these analyses it can be suggested that thestability of cathepsin L and other lysosomal cysteine proteinasesis a result of a delicate balance between hydrophobic and ionicinteractions. The results of this work are analogous to thefindings for cathepsins B and L at alkaline pH [23±25,38,44],indicating that the inactivation mechanisms at alkaline andacidic pH are similar. However, the number and nature of thecharged groups affected by the pH changes are different in thetwo pH regions. The inactivation mechanism can be representedby the following scheme (Scheme 2).

In this scheme, pH-induced inactivation of cysteine protein-ases starts in the active site region and is followed by domainseparation and further unfolding of the enzyme [25,44]. There isappreciable evidence that an intact active site is important andnecessary for the stability of the enzyme, in support of theproposed mechanism. Several active-site directed ligands, i.e.small synthetic substrates (Fig. 2) [23]; and competitive inhi-bitors [23,26], have been found to substantially stabilize, andeven to protect, cathepsin L against pH-induced denaturation.Furthermore, synthetic propeptides have been shown to bestrong inhibitors of cysteine proteinases, including cathepsin L[45±47]. From the crystal structures of procathepsin L [19]and procathepsin B [48,49], it is evident that the propeptide

Fig. 7. Sequence of human cathepsin L with main cathepsin D cleavage

positions indicated. All experimental details are given in the Materials and

Methods section. The most extensive cleavage was found between Cys98-

Lys99, with two other cleavage sites also being identified (Gly149-Ile150,

Gly191-Met192). The cleavage positions are indicated with arrows.

Scheme 2. The proposed mechanism for pH-induced inactivation of

papain-like cysteine proteases. -S± +HIm represents the active-site thiolate-

imidazolium ion pair, kinac is the inactivation rate constant, and E* is the

inactivated enzyme.

q FEBS 1999 pH regulation of cathepsin L (Eur. J. Biochem. 259) 931

completely covers the active site of the enzyme, therebypreventing access of substrate and solvent molecules to thissite. Such an interaction could also explain the unusual stabilityof procathepsins at neutral pH [5]. The active site Cys25 istherefore a likely candidate for one of the two groups with pKa

values <3.5, which were found to be involved in the rate-limiting steps of inactivation process. Furthermore, the pKa

value of the rat enzyme was found to be in a similar range(<3.3 ^ 0.2), supporting this idea [50].

At acidic pH, however, the propeptide of cysteine proteinasescan be cleaved, resulting in activation of the proenzymes.Cathepsin L, bound to negatively charged surfaces, was found tobe activated at pH 5.5±6.0, which is the pH of newly formedlysosomes [51]. With the maturation of lysosomes, their pH willdrop and can reach values as low as 3.8 in intact cells [52,53]. Atlower pH, all ligands, including substrate, inhibitors and pro-peptide, have been found to form only weak complexes withcathepsin L [26,46,54], indicating that the enzyme may beslowly, but irreversibly denatured as the lysosomes mature. Thedenatured cathepsin L could later be proteolytically degraded bycathepsin D, which is highly active at acidic pH [28], but latentat higher pH. The acid-induced inactivation and unfolding ofcathepsin L may thus be involved in the turnover of the enzymewithin lysosomes.

On the basis of these results, it is possible to suggest that thepH-induced inactivation followed by proteolytic degradationcould be a more general process, common to all lysosomalcysteine proteinases. Furthermore, the different pH-optima oflysosomal proteinases [1 29], spanning a wide range from pHclose to neutral for cathepsin H [1] to acid pH (3.0±4.0) forcathepsin D [28], suggest that the lysosomal proteinases areactive only at certain stages of the lysosome lifetime, acting in acascade-like manner with the pH gradient as the controllingfactor. Cathepsins B and S, although found to be more stable atneutral pH [38,55], were rapidly and irreversibly denatured atacid pH [56] (B. LenarcÏicÏ, unpublished results), in agreementwith the proposed hypothesis. By contrast, cathepsin D, a likelycandidate to degrade the acid-denatured proteinases, is known tobe stable at pH 3.5±4.0 [28], and show very little activity abovethis pH. Therefore, we suggest that cathepsin D is only involvedin the degradation of denatured proteinases in the most acidlysosomes, whereas the bulk of protein degradation is completedby cysteine proteinases prior to their denaturation in the lessacidic lysosomes. This is in agreement with the results of Saftiget al. [33]. As no endogeneous inhibitors of cathepsin D areknown, some other mechanisms for cathepsin D degradation andregulation of its activity at low pH must exist.

In conclusion, cathepsin L was found to be irreversiblyinactivated at acidic pH. The presence of small substrate mol-ecules substantially slowed down, but did not prevent theinactivation process, indicating that they stabilize the activeconformation of the enzyme. Inactivation was found to beaccompanied by pronounced structural changes, suggesting thatdenaturation is responsible for the activity loss. We suggest thatthe irreversible acid denaturation of lysosomal cysteine protein-ases could be a physiologically important mechanism for theregulation of their activities, and that cathepsin D may beinvolved in their physiological turnover.

ACKNOWLEDGEMENTS

This work was supported by the Swedish Medical Research Council (Project

no. 4212) and by the Ministry of Science of Republic of Slovenia. B.T. was

also supported by a FEBS Fellowship.

REFERENCES

1. Kirschke, H., Barrett, A.J. & Rawlings, N.D. (1995) Lysosomal

cysteine proteases. In Protein Profile, Vol. 2 (Sheterline, P., ed.),

pp. 1587±1643. Academic Press, London.

2. Barrett, A.J. & Kirschke, H. (1981) Cathepsin B, cathepsin H and

cathepsin L. Meth. Enzymol. 80, 535±561.

3. Maciewicz, R.A., Etherington, D.J., Kos, J. & Turk, V. (1987)

Collagenolytic cathepsins of rabbit spleen: a kinetic analysis of

collagen degradation and inhibition by chicken cystatin. Collagen

Relat. Res. 7, 295±304.

4. Mason, R.W., Wilcox, D., Wikstrom, P. & Shaw, E.N. (1989) The

identification of active forms of cysteine proteinases in Kirsten-virus-

transformed mouse fibroblast by use of a specific radiolabelled

inhibitor. Biochem. J. 257, 125±129.

5. Mason, R.W., Gal, S. & Gottesman, M.M. (1987) The identification of

the major excreted protein (MEP) from a transformed mouse

fibroblast cell line as a catalytically active precursor form of cathepsin

L. Biochem. J. 248, 449±454.

6. McCoy, K., Gal. S., Schwartz, R.H. & Gottesman, M.M. (1988) An acid

protease secreted by transformed cells interferes with antigen

processing. J. Cell. Biol. 106, 1879±1884.

7. Goretzki, L., Schmitt, M., Mann, K., Calvete, J., Chucholowski, N.,

Kramer, M., Guenzler, W.A., Jaenicke, F. & Graeff, H. (1992)

Effective activation of the proenzyme form of the urokinase-type

plasminogen activator (pro-uPA) by the cysteine protease cathepsin L.

FEBS Lett. 297, 112±118.

8. Kasai, M., Shirasawa, T., Kitamura, M., Ishido, K., Kominami, E. &

Hirokawa, K. (1993) Proenzyme from of cathepsin L produced by

thymic epithelial cells promotes proliferation of immature thymocytes

in the presence of IL-1, IL-7, and anti-CD3 antibody. Cell. Immunol.

150, 124±136.

9. Kasegawa, H., Tagami, K., Ohba, Y., Sumitani, K., Kawata, T. &

Katunuma, N. (1995) Secretion and processing mechanisms of

procathepsin L in bone resorption. FEBS Lett. 370, 78±82.

10. Dufour, E., Obled, A., Valin, C. & Bechet, D. (1987) Purification and

amino acid sequence of chicken liver cathepsin L. Biochemistry 26,

5689±5695.

11. Mason, R.W. (1986) Species variants of cathepsin L and their

immunological identification. Biochem. J. 240, 285±288.

12. Mason, R.W., Taylor, M.A.J. & Etherington, D.J. (1984) The

purification and properties of cathepsin L from rabbit liver. Biochem.

J. 217, 209±217.

13. Mason, R.W., Green, G.G. & Barrett, A.J. (1985) Human liver cathepsin

L. Biochem. J. 226, 233±241.

14. Towatari, T. & Katunuma, N. (1988) Amino acid sequence of rat liver

cathepsin L. FEBS Lett. 236, 57±61.

15. Wada, K., Takai, T. & Tanabe, T. (1987) Amino acid sequence of

chicken liver cathepsin L. Eur. J. Biochem. 167, 13±18.

16. Ritonja, A., PopovicÏ, T., Kotnik, M., Machleidt, W. & Turk, V. (1988)

Amino acid sequence of the human kidney cathepsins H and L. FEBS

Lett. 228, 341±345.

17. Ritonja, A., Coetzer, T.H.T., Pike, R.N. & Dennison, C. (1996) The

amino acid sequences, structure comparisons and inhibition kinetics of

sheep cathepsin L and sheep stefin B. Comp. Biochem. Physiol. B 114,

193±198.

18. Joseph, L.J., Chang, L.C., Stamenkovich, D. & Sukhatme, V.P. (1988)

Complete nucleotide and deduced amino acid sequences of human and

murine preprocathepsin L. An abundant transcript induced by

transformation of fibroblasts. J. Clin. Invest. 81, 1621±1629.

19. Coulombe, R., Grochulski, P., Sivaraman, J., Menard, R., Mort, J.S. &

Cygler, M. (1996) Structure of human procathepsin L reveals the

molecular basis of inhibition by the prosegment. EMBO J. 15,

5492±5503.

20. Fujishima, A., Imai, Y., Nomura, T., Fujisawa, Y., Yamamoto, Y. &

Sugawara, T. (1997) The crystal structure of human cathepsin L

complexed with E-64. FEBS Lett. 407, 47±50.

21. Gal, S. & Gottesman, M.M. (1988) Isolation and sequence of a cDNA

for human pro-(cathepsin L). Biochem. J. 253, 303±306.

22. Machleidt, W., Ritonja, A., PopovicÏ, T., Kotnik, M., Brzin, J., Turk, V.,

932 B. Turk et al. (Eur. J. Biochem. 259) q FEBS 1999

Machleidt, I. & MuÈller-Esterl, W. (1986) Human cathepsins B, H and

L: characterization by amino acid sequences and some kinetics of

inhibition by the kininogens. In Cysteine Proteinases and Their

Inhibitors (Turk, V., ed.), pp. 3±18.Walter de Gruyter, Berlin /New

York.

23. Turk, B., Dolenc, I., Turk, V. & Bieth, J.G. (1993) Kinetics of the

pH-induced inactivation of human cathepsin L. Biochemistry 32,

375±380.

24. Dehrmann, F.M., Coetzer, T.H., Pike, R.N. & Dennison, C. (1995)

Mature cathepsin L is substantially active in the ionic milieu of the

extracellular medium. Arch. Biochem. Biophys. 324, 93±98.

25. Turk, B., Bieth, J.G., BjoÈrk, I., Dolenc, I., Turk, D., Cimerman, N., Kos,

J., CÏ olicÏ, A., Stoka, V. & Turk, V. (1995) Regulation of the activity of

lysosomal cysteine proteinases by pH-induced inactivation and/or

endogenous protein inhibitors, cystatins. Biol. Chem. Hoppe-Seyler

376, 225±230.

26. Ogrinc, T., Dolenc, I., Ritonja, A. & Turk, V. (1993) Purification of the

complex of cathepsin L and the MHC class II-associated invariant

chain fragment from human kidney. FEBS Lett. 336, 555±559.

27. Bevec, T., Stoka, V., PungereÁieÁ, G., Dolenc, I. & Turk, V. (1996) Major

histocompatibility complex class II-associated p41 invariant chain

fragment is a strong inhibitor of lysosomal cathepsin L. J. Exp. Med.

183, 1331±1338.

28. Barrett, A.J. (1977) Cathepsin D and other carboxyl proteinases. In

Proteinases in Mammalian Cells and Tissues (Barrett, A. J., Ed.), pp.

209±248. Elsevier North-Holland Biomedical Press, Amsterdam-New

York-Oxford.

29. Mason, R.W. (1995) Lysosomal metabolism of proteins. In Biology of

the Lysosome, Subcellular Biochemistry, Vol. 27 (Lloyd, J.B. &

Mason, R.W., eds.), pp. 159±190. Plenum Press, New York±London.

30. Dingle, J.T., Poole, A.R., Lazarus, G.S. & Barrett, A.J. (1973)

Immunoinhibition of intracellular protein digestion in macrophages.

J. Exp. Med. 137, 1124±1141.

31. Mizuochi, T., Yee, S.T., Kasai, M., Kakiuchi, T., Muno, D. &

Kominami, E. (1994) Both cathepsin B and cathepsin D are necessary

for processing of ovalbumin as well as for degradation of class II

MHC invariant chain. Immunol. Lett. 43, 189±193.

32. Rochefort, H. (1990) Biological and clinical significance of cathepsin D

in breast cancer. Semin. Cancer Biol. 1, 153±160.

33. Saftig, P., Hetman, M., Schmahl, W., Weber, K., Heine, L., Mossmann,

H., KoÈster, A., Hess, B., Evers, M., von Figura, K. & Peters, C. (1995)

Mice deficient for the lysosomal proteinase cathepsin D exhibit

progressive atrophy of the intestinal mucosa and profound destruction

of lymphoid cells. EMBO J. 15, 3599±3608.

34. Turk, B., KrizÏaj, I., Kralj, B., Dolenc, I., PopovicÏ, T., Bieth, J.G. &

Turk, V. (1993) Bovine stefin C ± a new member of the stefin family.

J. Biol. Chem. 268, 7323±7329.

35. Smith, R. & Turk, V. (1974) Cathepsin D: rapid isolation by affinity

chromatography on haemoglobin-agarose resin. Eur. J. Biochem. 48,

245±254.

36. Pace, C.N., Vajdos, F., Fee, L., Grimsley, G. & Gray, T. (1995) How to

measure and predict the molar absorption coefficient of a protein.

Protein Sci. 4, 2411±2423.

37. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. (1951)

Protein measurement with Folin phenol reagent. J. Biol. Chem. 193,

265±275.

38. Turk, B., Dolenc, I., ZÏ erovnik, E., Turk, D., GubensÏek, F. & Turk, V.

(1994) Human cathepsin B is a metastable enzyme stabilized by

specific ionic interactions associated with the active site. Biochemistry

33, 14800±14806.

39. Tanford, C. (1968) Protein denaturation. Adv. Prot. Chem. 23, 121±282.

40. Tao, K., Stearns, N.A., Dong, J., Wu, Q. & Sahagian, G.G. (1994) The

proregion of cathepsin L is required for proper folding, stability, and

ER exit. Arch. Biochem. Biophys. 311, 19±27.

41. Maurel, P. (1978) Relevance of dielectric constant and solvent

hydrophobicity to the organic solvent effect in enzymology. J. Biol.

Chem. 253, 1677±1683.

42. Lestienne, P. & Bieth, J.G. (1980) Activation of human leukocyte

elastase activity by excess substrate, hydrophobic solvents, and ionic

strength. J. Biol. Chem. 255, 9289±9294.

43. Baici, A. & KnoÈpfel, M. (1986) Cysteine proteinases produced by

cultured rabbit V2 carcinoma cells and rabbit skin fibroblasts. Int. J.

Cancer 38, 753±761.

44. Dehrmann, F.M., Elliott, E. & Dennison, C. (1996) Reductive activation

markedly increases the stability of cathepsins B and L to extracellular

ionic conditions. Biol. Chem. Hoppe-Seyler 377, 391±394.

45. Fox, T., de Miguel, E., Mort, J.S. & Storer, A.C. (1992) Potent slow-

binding inhibition of cathepsin B by its propeptide. Biochemistry 31,

12571±12576.

46. Carmona, E., Dufour, E., Plouffe, C., Takebe, S., Mason, P., Mort, J.S.

& Menard, R. (1996) Potency and selectivity of the cathepsin L

propeptide as an inhibitor of cysteine proteases. Biochemistry 35,

8149±8157.

47. Taylor, M.A.J., Baker, K.C., Briggs, G.S., Connerton, I.F., Cummings,

N.J., Pratt, K.A., Revell, D.F., Freedman, R.B. & Goodenough, P.W.

(1995) Recombinant pro-regions from papain and papaya proteinase

IV-are selective high affinity inhibitors of the mature papaya enzymes.

Protein Eng. 8, 59±62.

48. Turk, D., Podobnik, M., Kuhelj, R., Dolinar, M. & Turk, V. (1996)

Crystal structures of human procathepsin B at 3.2 and 3.3 AÊ resolution

reveal an interaction motif between a papain-like cysteine protease

and its propeptide. FEBS Lett. 384, 211±214.

49. Cygler, M., Sivaraman, J., Grochulski, P., Coulombe, R., Storer, A.C. &

Mort, J.S. (1996) Structure of rat procathepsin B: model for inhibition

of cysteine protease activity by proregion. Structure 4, 405±416.

50. BroÈmme, D., Bonneau, P.R., Lachance, P., Wiederanders, B., Kirschke,

H., Peters, C., Thomas, D.Y., Storer, A.C. & Vernet, T. (1993)

Functional expression of human cathepsin S in Saccharomyces

cerevisiae. Purification and characterization of the recombinant

enzyme. J. Biol. Chem. 268, 4832±4838.

51. Mason, R.W. & Massey, S.D. (1992) Surface activation of pro-cathepsin

L. Biochem. Biophys. Res. Commun. 189, 1659±1666.

52. Berg, T., GjoÈen, T. & Bakke, O. (1995) Physiological functions of

endosomal proteolysis. Biochem. J. 307, 313±326.

53. Van Dyke, R.W. (1995) Acidification of Lysosomes and Endosomes. In

Biology of the Lysosome, Subcellular Biochemistry, Vol. 27 (Lloyd,

J.B. & Mason, R.W., eds.), pp. 331±360. Plenum Press, New York±

London.

54. Jerala, R., ZÏ erovnik, E., KidricÏ, J. & Turk, V. (1998) pH-induced

conformational transitions of the propeptide of human cathepsin L. A

role for a molten globule state in zymogen activation. J. Biol. Chem.

273, 11498±11504.

55. Kirschke, H., Wiederanders, B., BroÈmme, D. & Rinne, A. (1989)

Cathepsin S from bovine spleen. Biochem. J. 264, 467±473.

56. Stoka, V. (1997) Biochemical characterization of cathepsin S and

cruzipain: two related enzymes of the papain superfamily. PhD Thesis,

University of Ljubljana.